Lung and esophageal cancer represent common malignancies which cause the death of >180,000 individuals each year in the United States alone. Currently available chemotherapy both as neoadjuvant treatment and as palliative therapy for advanced disease has had a small but measurable effect on the survival of patients with these diseases. However, while a number of different drugs (for example alkylating agents, platinum compounds, 5FU, and taxanes) are available, no means currently exists for selecting therapy based upon the biology of a patient's individual tumor. We have shown, previously, that DNA hypermethylation associated promoter region silencing of the DNA damage repair gene O6-MGMT, can predict patients sensitive to alkylating agents, and these patients benefit with long-term survival. We now have preliminary data in esophageal cancer patients suggesting that hypermethylation of the checkpoint with forkhead and ring finger domains gene (CHFR) detected by Methylation Specific PCR (MSP) correlates with response to therapy with taxanes and results in prolonged survival of individual patients with CHFR methylation. In the R21 portion of this proposal, we intend to (1) complete a single institution, retrospective trial of neoadjuvant taxane-based therapy in esophageal cancer and extend our analysis to lung cancer patients who also have been treated with neoadjuvant taxanes;(2) detect methylation of CHFR in plasma from patients with esophageal cancer to determine if plasma can be used when tumor tissue is not available;(3) determine the performance characteristics of this assay and compare results between gel-based MSP and real-time MSP for the CHFR gene. In the R33 component of this application, we will validate both our gel-based and real-time tests for CHFR methylation in multi-institutional retrospective studies in lung and esophageal cancer. Then, we intend to apply mainly our real-time assay to prospective studies both in single and multi-institutional settings. These studies, for both neoadjuvant and palliative treatments, will determine the clinical utility of CHFR methylation to predict sensitivity of lung and esophageal cancer to taxanes. They may also provide the basis for targeted use of these agents in other cancer types, including colon, gastric, and head and neck cancer, where CHFR methylation is also frequently observed. Taxanes are a class of chemotherapeutic agents widely used to treat lung and esophageal cancer patients, but many patients do not respond to these drugs. We will further develop a test for CHFR methylation and determine whether CHFR methylation can be used to predict which patients will respond to taxane based chemotherapies.